Businesses seeking to take the friction out of operational decision-making are turning data streams into continuous intelligence.
No matter the industry in which they operate, all businesses strive to stay relevant; in the data-driven era of IoT, staying relevant means staying current. But surprisingly, many organizations still place their faith in historical databases over ongoing data. Such organizations will necessarily lack the information architecture they need to make informed decisions in rapidly changing environments. What they need, in a word (or two) is continuous intelligence (CI) — an approach to data-driven decision making that leverages technology for improved business outcomes.
Gartner predicts that by 2022, more than half of major new business systems will in some way exploit continuous intelligence capabilities. Because continuous intelligence is about better decision-making, successful implementation relies on improved data processing, AI, and/or machine learning capabilities. Ultimately, the right software platforms, that deliver the right data at the right time (and in the right way) can enable continuous intelligence along with its attendant benefits.
As the universe of applications for IoT and big data analytics have proliferated across the enterprise, concepts like “continuous intelligence” have been coined to guide the understanding and implementation of systems designed to activate all this new information. To that end, it’s worth differentiating continuous intelligence from similar ideas that are related, but not quite the same.
In a sense, continuous intelligence is more than the sum of its parts. A business can have multiple streams of real-time data in place alongside historical data, but “data” and “IoT” alone don’t define CI. Neither does the term refer exclusively to the speed at which data is produced or consumed. It does not describe the rate of information generation and transfer, nor does it denote a high throughput, even though these qualities are crucial to its success. Lastly, CI is neither defined by nor reliant upon the application of artificial intelligence.
CI is also distinct from “business intelligence” in the sense that business intelligence typically doesn’t move at a real-time pace — BI typically runs through slower, human-driven decision-making processes like meetings, committees, and reports. While AI may represent a core component of an organization’s business intelligence arm, it’s not typically designed to actuate decisions against real-time data. As a concept, business intelligence doesn’t refer to insights based on up-to-date information, nor does it imply ongoing action, even if it is used to inform bigger-picture decisions. BI is, fundamentally, periodic.
Continuous intelligence uses real-time data, automated or semi-automated processes, and intelligent information systems to augment human decision-making on an ongoing basis — in other words, any time a decision is made. CI systems enable “frictionless,” augmented analytics designed to inform human decisions with the most accurate data possible. By definition, CI should be embedded in the core of all business operations. It’s not a once-quarterly process for getting back on track or adjusting strategic direction — it should be an inherent part of how a business works and runs each and every minute of every day.
Continuous intelligence is, by definition, both real-time and “continuous.” Good CI relies on current data, as well as human inputs around observed events or situational touchstones. CI systems should be able to integrate a vast pool of divergent data sources at extremely high volumes. Applications used for continuous intelligence will necessarily tap into historical databases to some extent, but the fact that data offloaded to storage goes through duplications and security protocols can (somewhat ironically) make it less usable for fast decision automation. Instead, the priority data for CI tools is always the data generated in real-time.
In certain scenarios, continuous intelligence can benefit from AI and machine learning — capabilities which can allow for improved (read: superhuman) event stream processing. With CI, augmented analytics can effectively predict potentialities and prescribe optimal responses before human decision makers could even begin to evaluate post-mortem data. Continuous intelligence is designed for an ever-changing environment — and while shifting factors may make it harder for people to see a situation objectively, CI can not only adjust to such changes — it actually thrives on them.
Depending on the context, continuous intelligence implies partial decision automation. In discrete contexts wherein technologies are capable of fully governing certain operations — as with certain machinery or software — closed-loop automation through CI makes sense. But for the many organizations that rely on operational human action, CI’s role is more of a supporting one — removing the biases and mitigating the computational limitations of the human mind.
Continuous intelligence can be transformative for business operations. Data doesn’t have to disappear into vast databases — it can (and should) inform the most pressing organizational decisions. At the same time, automation (and artificial intelligence) can relieve humans of the burdens associated with the responsibility of making a “final” decision. In one sense, many of us already rely on technology to make automated decisions for us — for instance, when Waze reroutes you to avoid a traffic jam. Unsurprisingly, when applied to mammoth operational decisions, the potential impact is difficult to imagine. To use the transportation and logistics industry as an example, imagine shaving off a few minutes from every route, or avoiding extra deadhead miles or snagging an extra load on an existing delivery. These types of incremental, real-time decisions have a huge impact on the bottom-line. Not making real-time data-driven decisions drives poor customer experience, limits resource capacity and discourages employees from making the best decision possible due to a lack of data access.
Even more crucially, the “point” of continuous intelligence is that it is outcome-focused. The “improvement” in human decision-making refers to a quantifiable impact on business operations. With a CI mindset, organizations can get more value from their data as their operations become more effective. For instance, CI can help a manufacturer reduce downtime for their equipment with predictive analytics that pinpoints exactly when certain machines need attention. Condition-based maintenance means equipment is fixed only after it’s required. The implications go far beyond maintenance itself.
Although continuous intelligence can help remove some degree of human bias from the decision-making equation by surfacing hard data, a CI architecture can (and must) still be flexible and intuitive enough for human use. While some businesses may use CI to manage “self-regulating” machinery or software, all organizations ultimately run on human actions.
From emergency services to logistics to maintenance to security, organizations don’t need total “automation” — what they can benefit from, however, is less friction around the decision-making process. In the context of the OODA Loop ( “observe, orient, decide, act”), CI can help teams make better decisions faster, for more successful outcomes.
Ultimately, continuous intelligence is just one (valuable) component of situational awareness. Deploying intelligent applications which take advantage of real-time data is what many businesses need to overcome blind spots and bottlenecks, and finally bake informed decision-making into their operational strategy.
Of course, not all applications of CI are created equal — nor indeed are many of these altogether worthy of the moniker. Many may suffer from a lack of extensibility, have trouble integrating and activating critical information, or fail to successfully align with business-specific decision models.
Organizations must ensure that their CI solutions contribute to a common operational picture built on a foundation of real-time data derived from mission-critical assets (i.e. IoT devices). CI solutions must be contextual — in other words, crafted in support of any given organization’s unique objectives. Achieving this contextuality requires platforms that balance the need for powerful built-in capabilities with the need for flexibility and customizability. Continuous intelligence is neither a concept untethered from real applications nor a rigid technological framework — it is a self-evidently valuable and imminently achievable operational asset that, as a matter of course, will take different forms from one use case to the next.